The suitability of an uncemented hydroxyapatite coated (HAC) hip hemiarthroplasty stem for intra-capsular femoral neck fractures in osteoporotic elderly patients: the Metaphyseal-Diaphyseal index, a solution to preventing intra-operative periprosthetic fracture
1 Department of Trauma & Orthopaedics, Queen Elizabeth Hospital, South London Healthcare NHS Trust, Stadium Road, Greenwich, London, SE18 4QH, UK
2 Department of Trauma & Orthopaedics, Princess Royal University Hospital, South London, Healthcare NHS Trust, Farnborough Common, Orpington, BR6 8ND, Kent, UK
3 Department of Trauma & Orthopaedics, Darent Valley Hospital, Darenth Wood Road, Dartford, DA2 8DA, Kent, UK
4 Institute of Postgraduate Medicine,Brighton & Sussex Medical School, University of Sussex, Brighton, East Sussex, BN1 9PX, UK
Journal of Orthopaedic Surgery and Research 2011, 6:59 doi:10.1186/1749-799X-6-59Published: 18 November 2011
This study will seek to identify a measurable radiographic index, the Metaphyseal-Diaphyseal Index (MDI) score to determine whether intra-operative fracture in osteoporotic bone can be predicted.
A 5 year prospective cohort of 560 consecutive patients, undergoing hemiarthroplasty (cemented or uncemented), was evaluated. A nested case-control study to determine risk factors affecting intra-operative fracture was carried out.
The Vancouver Classification was used to classify periprosthetic fracture.
The MDI score was calculated using radiographs from the uncemented group. As a control (gold standard), Yeung et al's Canal Bone Ratio (CBR) score was also calculated. From this, a receiver operating characteristic (ROC) curve was formulated for both scores and area under the curve (AUC) compared. Intra and inter-observer correlations were determined.
Cost analysis was also worked out for adverse outcomes.
Four hundred and seven uncemented and one hundred and fifty-three cemented stems were implanted. The use of uncemented implants was the main risk factor for intra-operative periprosthetic fracture.
Sixty-two periprosthetic fractures occurred in the uncemented group (15.2%), nine occurred in the cemented group (5.9%), P < 0.001. The revision rate for sustaining a periprosthetic fracture (uncemented group) was 17.7%, P < 0.001 and 90 day mortality 19.7%, P < 0.03.
MDI's AUC was 0.985 compared to CBR's 0.948, P < 0.001. The MDI score cut-off to predict fracture was 21, sensitivity 98.3%, specificity 99.8%, positive predictive value 90.5% and negative predictive value 98%. Multivariate regression analysis ruled out any other confounding factors as being significant.
The intra and inter-observer Pearson correlation scores were r = 0.99, P < 0.001.
JRI uncemented hemiarthroplasty has a significantly higher intra-operative fracture rate. We recommend cemented arthroplasty for hip fractures. We propose a radiographic system that may allow surgeons to select patients who are good candidates for uncemented arthroplasty, but it needs prospective validation.